There is an outstanding problem in artificial intelligence of how to apply commonsense knowledge to transform domain-specific knowledge so as to adapt it to current problem-solving needs as defined by a supplied context. Expert systems are currently used for the application of domain-specific knowledge for the solution of complex problems. However, they are brittle, cannot get too large or they become too costly to maintain, and cannot effectively apply commonsense reasoning.